1. Arize AX
  2. Phoenix OSS
  3. Docs
  4. Learn
  5. Company
  6. Pricing
Book a demo Sign in Sign up

Blog

Filter
close Clear all
Rise of the AI engineer: Why AI evals fail before the evaluation begins

Hamel Husain explains why AI evals fail before the evaluation begins

Hamel Husain explains why ambiguous inputs, generic metrics, and disconnected review workflows can make AI evaluations misleading, and how developers can build a better process around real production data.

6 minutes read
By Sara Verdi | 6 minutes read
Agent Engineering Agent Evaluation Agents AI Engineering AI Evaluation LLM Evals LLM Evaluation

How to improve agent skills with tracing and evals

A skill cut agent costs by 44% and latency by 56%, but also reduced answer completeness. Here’s how tracing, evals, and a long-running agent exposed and corrected the regression.

7 minutes read
By Yusuf Cattaneo | 7 minutes read
Agent Engineering Agent Evaluation Agent Observability Agents AI Engineering AI Evaluation AI Observability LLM Evals LLM Evaluation

How to write effective AI agent skills: 6 data-backed practices

Three recent studies show what actually makes an AI agent skill effective: human expertise, compact procedures, tight routing, harness-specific testing, and eval-gated changes—not longer Markdown.

10 minutes read
By Laurie Voss | 10 minutes read
Agent Engineering Agent Evaluation Agents AI Engineering AI Evaluation LLM Evals Research
Kappa, Alpha, or F1? A practical guide to measuring human–LLM judge alignment

How to measure human-LLM judge alignment

No single metric proves an LLM judge is trustworthy. This field guide shows how to measure human–human agreement, compare it to LLM–human agreement, and diagnose errors with precision, recall, and F1.

14 minutes read
By Elizabeth Hutton | 14 minutes read
LLM As A Judge LLM Evals LLM Evaluation

How do you make an LLM, anyway? Microsoft just published a textbook.

Microsoft published a 109-page technical report on MAI-Thinking-1. Here’s the abbreviated version of how a modern lab actually trains a frontier reasoning model — from scraping the web to reinforcement learning, judges, and anti-cheating.

10 minutes read
By Laurie Voss | 10 minutes read
AI Evaluation LLM As A Judge LLM Evals LLM Evaluation
AI evals are a data science problem — Hamel Husain at Arize Observe 2026

AI evals are a data science problem: What most teams get wrong

Hamel Husain explains why the best AI teams treat LLM judges like classifiers, not dashboards.

9 minutes read
By Sara Verdi | 9 minutes read
AI Evaluation LLM As A Judge LLM Evals

Meet PXI: the AI engineering agent inside Phoenix

An AI engineering agent built into Phoenix. It works like a coding agent, just point it at your telemetry instead of a source tree.

16 minutes read
By Mikyo King | 16 minutes read
Agent Engineering Agent Evaluation Agents AI Engineering AI Evaluation AI Observability LLM Evals LLM Observability Open Source Product Releases
A graphic that shows Arize Phoenix and a headline that says

How to ship a local LLM that matches frontier LLMs with evals and prompt engineering

Most production AI features don’t need a frontier model. Here’s how capability evals and prompt engineering can help ship a local SLM that matches frontier-model quality with lower latency and cost.

14 minutes read
By RL Nabors | 14 minutes read
AI Evaluation LLM Evals LLM Evaluation Prompt Engineering
Load More

Subscribe to our resources and blogs

Subscribe